الفهرس 
TitleOnly 14 pages are availabe for public view
 |  | from | 130 |  |  |
| | | from | 130 | | |
Abstract
Thyroid nodules represent the most common disorder of the endocrine system. Making a precise distinction between benign and malignant is essential so that patients with malignancy receive definitive treatment and patients with benign nodules avoid unnecessary surgery.
The approach to thyroid nodule assessment is multidisciplinary, integrating clinical, radiological, cytological and molecular analysis. US-Guided FNAC is considered the world-wide gold standard method for diagnosing thyroid nodules. Its results are useful for malignancy risk stratification. The Bethesda System for Reporting Thyroid Cytopathology is a 6-category classification system fashioned to standardize the interpretation of thyroid cytology.
Follicular neoplasm (Bethesda category IV) encompasses a wide spectrum of diseases ranging from benign to malignant follicular derived thyroid lesions. They are characterized by their overlapping cytomorphologic features, so can’t be accurately distinguished by FNA alone.
Even with its countless advantages, sometimes FNA does not yield sufficient information for precise diagnosis and so the risk of false negative or intermediate diagnoses always exists, therefore complementary use of cell block is preferred.
Cell block is considered true microbiopsy and is known by its similarity to histology, representing an interface between cytology and histopathology. It permits the identification of architectural patterns same as those observed in histological sections, which when associated with morphological cellular details present in the cytological preparations, facilitates a definitive diagnosis. Furthermore, CB can be used for ancillary techniques as immunohistochemistry.
CD56 has high expression rates in normal thyroid tissue and several follicular lesions, but it is usually lost in PTC.
HBME-1 is not expressed in normal thyroid tissue. On the contrary, HBME-1 was reported to be overexpressed in most PTCs, and also in a fraction of FTCs.
The present study aims at evaluating the diagnostic utility of CD56 and HBME- 1 performed on cell block of thyroid aspirates of cases categorized cytologically as Bethesda IV and correlating these results with the final histopathologic diagnosis in an attempt to test their value as preoperative markers to differentiate benign from malignant thyroid lesions, therefore narrowing the spectrum of this Bethesda group.
To achieve this aim, the present study was undertaken on 50 prospective cases diagnosed as Follicular neoplasm/Suspicious of follicular neoplasm (FN/SFN) (Bethesda category IV) by FNAC, together with their prepared cell block and the subsequently processed thyroidectomy specimens. Cases were collected from the pathology department, Medical Research Institute, Alexandria University in the period between December 2018 and July 2019.
Summary, Conclusion and Recommendations
91
The 50 studied cases were diagnosed on thyroidectomy basis and classified into 32 benign cases (64%) and 18 malignant cases (36%). The 32 benign cases included: 22 cases of follicular adenoma (44%), 4 cases of hyperplastic nodule (8%), 5 cases of Hürthle cell adenoma (10%) and 1 case of Hashimoto’s thyroiditis (2%). The 18 malignant cases included: 5 cases of follicular carcinoma (10%), 9 cases of PTC [7 cases of FVPTC (14%), 1 case of classic PTC (2%) and 1 case of oncocytic variant of PTC (2%)], 1 case of medullary thyroid carcinoma (2%) and 3 cases of Hürthle cell carcinoma (6%).
Patients’ age ranged from 20-73 years with a mean of age 43.82 ± 13.674 years. Thirty eight cases were females (76%) while twelve cases were males (24%). No statistical significance was found between BTLs and MTLs regarding age or sex. Nodule size was <2 cm in 14 cases (28%), from 2 to 4 cm in 24 cases (48%) and >4 cm in 12 cases (24%). A significant correlation was noted between nodule size in BTLs and MTLs (p value= 0.005).
Cases were subgrouped into “Favoring FNs” and “Indeterminate FNs” and were examined in terms of both cytological architecture and nuclear characteristics. Regarding architecture features, trabecular pattern and branching mono-layered sheets showed high statistical significance in differentiating “Favoring FNs” and “Indeterminate FNs” subgroups (p value= 0.021 & <0.001 respectively). As for nuclear features, only chromatin clearing showed a significant correlation (p value= 0.009).
In the present work, usage of cell block in adjunct to smears showed improvement in cellularity, allowed better assessment of architectural pattern and decreased background obscuring material.
CD56 in the current study showed expression in 30 (93.8%) out of 32 BTLs, while it was positive in 7 (38.9%) out of 18 MTLs, showing high significant differentiation between immunohistochemical expression of CD56 in BTLs & MTLs (p value <0.001). Nevertheless, CD56 showed no statistical significant association in expression among different types of BTLs (p value= 0.523) as well as among different MTLs (p value= 0.199).
Conversely, HBME-1 was expressed more in MTLs, being positive in 13 (72.2%) out of 18 cases, while HBME-1 was expressed in 6 (18.7%) out of 32 BTLs, showing high significant association between immunohistochemical expression of HBME-1 in BTLs & MTLs (p value <0.001). HBME-1 showed no statistical significant association in expression among different BTLs (p value= 0.243) as well as among different types of MTLs (p value= 0.679).
The sensitivity of CD56 in differentiating BTLs from MTLs was low (61%), specificity was very high (94%), PPV was 85%, NPV was 81% and DA was 82%. On the other hand, HBME-1 showed higher sensitivity (72%) and lower specificity (81%), while PPV was 68%, NPV was 84% and DA was 78%.
As stated earlier, CD56 was expressed more in benign thyroid lesions (BTLs) (AUC= 0.844), whereas HBME-1 was expressed more in malignant thyroid lesions (MTLs) (AUC= 0.819), concluding that each marker when used separately can differentiate BTLs from MTLs with high statistical significance (each showed p value <0.001). Combining both markers showed greater discriminative ability in differentiation (AUC= 0.914).
Both CD56 and HBME-1 were able to differentiate follicular variant of papillary thyroid carcinoma (FVPTC) from follicular adenoma (FA) when used separately with high statistical significance (AUC= 0.938 & 0.919 respectively) (p value of each =0.001). Combining both markers showed greater discriminative ability (AUC= 0.961) (p value <0.001).
Neither CD56 nor HBME-1 was effective in differentiating follicular carcinoma (FC) from FA (AUC=0.591 & 0.755 respectively) (p value=0.533 & 0.081). Nevertheless, combining both markers could differentiate with statistical significance (AUC=0.891) (p value=0.007).
CD56 showed ability in differentiating FC from FVPTC with statistical significance (AUC=0.871) (p value= 0.035). On the contrary, HBME-1 couldn’t differentiate between the two neoplasms (AUC=0.829) (p value= 0.062). Combining both markers showed higher statistical significance in differentiation than that of CD56 (AUC=0.900) (p value=0.023).
Conclusion
In cytological smears, trabecular pattern proved to be significantly present in “Favoring FN” category. On the other hand, both branching monolayered sheets as well as chromatin clearing proved to be significant features in “Indeterminate FN” category
Nuclear features seen in papillary thyroid carcinoma as nuclear clearing, grooving and intranuclear inclusions alone are not strong findings to diagnose PTC cytologically, since these features can be encountered in some cases of follicular adenoma as well as follicular carcinoma.
Cell block in conjunction with FNAC has refined the diagnosis of Bethesda IV lesions, since the former provided higher cellularity and better architectural pattern assessment, along clean appearance of the background.
Another advantage of cell block in the present study was its diagnostic usefulness in applying immunostaining for both CD56 and HBME-1, resulting in narrowing the spectrum of Bethesda IV category.
CD56 immunostaining in the present study proved significantly effective in differentiating benign and malignant thyroid lesions (p<0.001). Nevertheless, it didn’t prove to be significant in differentiating various benign thyroid lesions (p= 0.523) as well as various malignant thyroid lesions (p=0.199).
HBME-1 immunostaining in the present study proved significantly effective in differentiating benign and malignant thyroid lesions (p<0.001). However, it didn’t prove to be significant in differentiating various benign thyroid lesions (p=0.243) as well as between different malignant thyroid lesions (p=0.679).
In comparing both CD56 and HBME-1, CD56 gave low sensitivity (61%) and high specificity (94%), while HBME-1 showed higher sensitivity (72%) and lower specificity (81%).
The combination of both markers in the present study raised the significant value in differentiating benign from malignant thyroid lesions reflected by the high statistical value of AUC (=0.914).
In the present study, a promising significant result was found for both CD56 and HBME-1, being able to differentiate follicular variant of papillary carcinoma from
Summary, Conclusion and Recommendations
93
follicular adenoma (p of each= 0.001). On combining both markers, our result gave a greater significant value in the above differentiation. (p<0.001).
Each CD56 and HBME-1 were insignificant in differentiating follicular carcinoma from follicular adenoma (p value =0.533 and 0.081 respectively). However, combining both markers gave significant high statistical value in such differentiation (p=0.007).
In contrast to HBME-1, CD56 immunostaining could significantly differentiate follicular carcinoma from follicular variant of PTC (p=0.035), while combining both markers gave a higher significant value in the differentiation (p=0.023).
Recommendations
Applying the current study on a wider scale of Bethesda IV lesions is recommended in a trial to subcategorize the Bethesda IV group into three entities including a. Favoring benign, b. Borderline and c. Favoring malignant to convey a near correct impression of the pre-operative diagnosis, subsequently affecting the efficacy of the management. Such subcategorization can be followed similar to the Japanese cytology reporting system.
In approaching each case of thyroid nodule, it is highly advised to perform both FNAC and cell block as a routine procedure, to get all needed cytological information in addition to the precious tissue yield of cell block, being valuable in applying immunostaining whenever needed.
Since cell block gave precious information together with FNAC, it is recommended in the sampling procedure to give a separate aspiration pass for the cell block, apart from that of the FNAC, instead of considering the residual material of FNAC pass as cell block.
The application of both CD56 and HBME-1 on cell block of all Bethesda IV cases is recommended in order to refine the pathological diagnosis and narrow the spectrum of the differential diagnosis.
Moreover, both CD56 and HBME-1 need to be further studied on the indeterminate thyroid nodules (Bethesda III, IV and V) to evaluate their efficacy in differentiating benign from malignant cases.
Expansion of this work needs to be applied using a wider panel of immunostaining markers as CK19, Galectin 3 and TROP-2 in concomitance to the combination of CD56 and HBME-1 in the present study, in order to determine the most superior combination for the diagnostic approach of various thyroid lesions.
To study the application of both CD56 and HBME-1 in the diagnosis of “follicular-patterned neoplasms with borderline clinical behaviour”, including follicular tumour and well differentiated tumour of uncertain malignant potential as well as non-invasive follicular tumour with papillary like nuclear features (NIFTP).
The application of recent molecular techniques as polymerase chain reaction and in situ hybridization on tissue-rich cell block needs to be evaluated to refine Bethesda IV category.
Multidisciplinary approach is essential for proper diagnosis of thyroid nodules.